Gas transport properties in waterborne polymer nanocomposite coatings containing organomodified clays.
Keywords Nanocomposite coating, Clay, Exfoliation, Gas diffusion, Solubility, Permeability
Polymer/clay nanocomposites with enhanced barrier properties
The incorporation of layered aluminosilicate particles into polymer matrices is found to be effective in improving the physical, mechanical, and thermal properties of polymers in comparison with conventional composites. (1-3) In addition, when the inorganic phase is well dispersed in the organic host, the gas barrier properties of the resulting nanocomposite can be enhanced up to several orders of magnitude. (4,5) This is due to the exfoliated crystalline layers acting as impermeable barriers to the diffusing gas penetrant, forcing its molecules to follow a long tortuous diffusion path. The clay delays, thus, their passage through the material. (6) This type of barrier is often called passive in contrast to a reactive barrier, in which the solute reacts with some ingredients of the host matrix.
The efficiency of passive inorganic particles to improve the gas barrier properties of the final material is directly related to their form, their aspect ratio, the volume fraction and orientation. (7,8) These factors influence the tortuosity, which is a measure of the diffusion path complexity. When the tortuosity increases, the gas molecules are forced to find alternative longer pathways to diffuse through the nanocomposite. As the diffusion path increases and the required time to travel through the film is prolonged, the overall rate of diffusion is reduced. In order for the nanolayers to contribute efficiently to the permeation properties, they should be exfoliated, uniformly dispersed into the matrix, and oriented perpendicular to the diffusion direction. It should be noted that as their volume fraction increases, they have a tendency to agglomerate.
The aim of the present article is to develop efficient methods to prepare coatings with enhanced barrier properties. The polymer/clay nanocomposite coatings consist of a polymeric matrix and montmorillonite (MMT) clay reinforcement. The phylosilicates, which include the bentonite and the MMT, are widely available and inexpensive minerals. Waterborne resin emulsions are environmentally friendly and economic coatings and are used here as the organic matrix. Further, since the neat clay is hydrophilic, the water in these coatings can be used as the suspension medium both for the resin colloidal droplets ("particles") and the exfoliated clay. The efficiency of this method to prepare coatings with enhanced barrier properties has been demonstrated (9) for high aspect ratio hectorite dispersed in acrylic-based waterborne coating. However, at the final stage of the formation of the continuous coating film, the water has evaporated and the neat hydrophobic resin comes into contact with the clay. It is not clear, therefore, whether untreated hydrophilic clay gives the optimum properties to the coating.
Two common resin systems that are used for waterborne coatings are acrylic-based and polyurethane-based. These two organic resins differ substantially in their degree of hydrophobicity, even though their particles in the coating are highly surface-treated to form the stabilized suspension in water. Their different polarity, therefore, may cause differences in their interactions with the clay at the final stages of film formation. Further, the clay can also be treated to become less hydrophilic by surface modification, e.g., using organic ammonium cations. (10) Thus, there is a range of organomodified clays available with different degrees of hydrophilicity, which may react differently with resins with different hydrophobicity in their neat state, such as the PU-based and the acrylic-based resins used here.
The organomodified clays, however, may not exfoliate spontaneously in the waterborne coating for the preparation of the nanocomposite. An organic co-solvent can be used, then, to form the suspension medium. The present article examines the parameters that affect the interactions between the different components used during the preparation of nanocomposite coatings with high barrier properties. The parameters that have an impact on the final nanostructure (exfoliated or not) are identified and their effect on the transport properties (gas diffusivity) of the coating are discussed.
Film formation of a waterborne nanocomposite coating
During film formation of a waterborne coating, the aqueous colloidal dispersion of monomer droplets ("resin particles" in the coatings literature) transforms into a continuous and homogeneous cured film. (10,11) The entire process goes through various overlapping steps:
(1) The water and/or the solvent evaporate and the particles become ordered.
(2) The colloidal particles deform and coalesce and the boundaries disappear.
(3) A crosslinking reaction takes place and the continuous film (the coating) is formed on the substrate.
The composite coatings developed here consist of a thermosetting polymer matrix and MMT clay reinforcement. Since a waterborne resin dispersion is used for the matrix, the addition of the nanoclay complicates the procedure of film formation and includes the following processes (12,13):
(1) The clay needs to be dispersed into the suspension medium and exfoliated.
(2) The whole suspension needs to become macroscopically homogeneous, i.e., the dispersion of resin particles and exfoliated clay nanoplatelets in the water needs to be kept spatially homogeneous.
(3) The water has to evaporate for the resin to be cured. This would eventually bring the clay nanoplatelets, which are initially suspended in the water, into a suspension within the continuous polymeric (organic) matrix. In order to obtain a coating with good barrier properties, the nanoclay should remain exfoliated during this process.
This procedure is not simple and it is difficult to guarantee that the clay will not collapse during its transfer from the aqueous into the organic phase or during curing. In the present work, we examined several preparation methods that may lead to this transfer without losing the final exfoliated state of the clay.
One preparation method employed the direct dispersion of hydrophilic platelets (MMT) into the stabilized aqueous dispersion of the uncured resin. This was followed by the evaporation of the water and the crosslinking reaction of the matrix which gave the final nanocomposite coating. In another method, organically modified MMT particles were dispersed in an organic solvent, and the lot was mixed with the waterborne matrix and let to dry and cure. Each method sets different requirements for the components used for the preparation of the composite. Thus, the two resin formulae, which show different hydrophobicity in their neat state, may result in products with different properties when combined with hydrophilic or organomodified MMT clays.
The drying process during the formation of the coating is complicated in the present case by three principal effects: (i) The strong dependence of solvent diffusivity on solvent content; (ii) the plasticizing effect of the solvent; and (iii) the viscoelastic memory effects of the polymer matrix when it enters the glassy state. The rate of the drying process is limited by heat transfer and solvent evaporation during the initial stage, and by diffusion and chain relaxation at longer times. If the solvent transport within the film is solely governed by diffusion, the kinetics is called Fickian. As the polymer concentration increases during drying, the chain mobility decreases and solvent diffusion through the matrix slows down. This effect is more prominent when the film enters the glassy state either during drying or during curing, where the chains become frozen.
In the glassy state, the rate of contraction of the matrix volume during drying may be slower than the curing time scale. The solvent acts as a plasticizer modifying the transport coefficients of the polymer matrix and its viscoelastic relaxation. A special Deborah number, [De.sub.D] = [[tau].sub.M]/[[tau].sub.D], is frequently used to assess the relative importance of solvent diffusion and structure relaxation, where [[tau].sub.M] is the structure relaxation time and [[tau].sub.D] is the characteristic time for diffusion, [[tau].sub.D] = [L.sup.2]/D, with L the sample thickness and D the mutual diffusion coefficient. For very high values of this Deborah number (thin films or slow relaxation), the long-time sorption dynamics are governed by viscoelasticity and memory effects, whereas for low De (thick film or fast relaxation) diffusion dominates. At the initial stages of drying, this relaxation time is low due to the plastification of the solvent. Later, as the solvent leaves the coating surface, the value of De becomes 0(1) and nonFickian diffusion can become significant. (14)
When the exfoliated clay has been dispersed in the waterborne coating before drying, then the most probable positioning of the clay particles will be between the resin particles. Thus, when drying proceeds, the clay particles become entrapped between these droplets. Hopefully, they remain there during the coagulation of the resin particles and the curing of the matrix and form the dispersed clay phase in the final nanocomposite.
The proper combination of solvent(s)/dispersion medium and a controlled evaporation rate can result in the successful entrapping of the nanoparticles between the resin droplets, resulting in a well-exfoliated nanocomposite. (15) The combination of slow solvent evaporation rates and curing temperatures set below the [T.sub.g] of the final resin provides a good chance to entrap the delaminated inorganic nano-objects between the contact surfaces of the polymer droplets and hinder their reaggregation during the coalescence stage. (16,17) On the contrary, fast evaporation of the solvent could result in reaggregation of the clay and the formation of defects, such as bubbles or pores, which are detrimental for the barrier properties of the film. (18)
Two organic waterborne resin systems were used as matrix precursors: a polyurethane and an acrylic polymer emulsion, provided by Nuplex Resins (Bergen op Zoom, the Netherlands), noted here as PU098 (PU) and Setalux 6768 (Sx), respectively. The colloidal droplets (60-70 nm in diameter) were made to have a hydrophobic core and a stabilizing hydrophilic shell, as shown in Fig. 1. The particles form colloidal dispersions of 40 vol% in demineralized water in both systems. Setalux 6768 is a self-crosslinkable resin at room temperature, whereas PU098 uses a methylated melamine-formaldehyde crosslinking agent (Cymel 328 by CYTEC Industries) and needs to be conditioned in a drying oven.
MMT clay was the dispersed phase in all nanocoatings. This clay is hydrophilic, i.e., incompatible with the hydrophobic polymers. Due to this incompatibility, we used natural MMTs modified with organic modifiers ("organomodified montmorillonite," oMMT), under the trade name of Cloisite[R] C10A, C25A, C93A, in addition to an octadecylamine-modified bentonite (OMBE). The chemical structures of the modifying ions are shown in Table 1. Bentonite Nanocor[R] was used as "untreated" MMT clay (BT) for comparison. The nanoclay powders were supplied from Sigma-Aldrich and Southern Clay Products (Texas, USA) and used as received.
In addition to demineralized water, two organic solvents were used as dispersion/suspension media for the preparation of the initial solution. Ethyl alcohol and acetone were chosen due to their miscibility with water, since both resins are used as aqueous dispersions.
Three groups of samples were prepared: (i) hydrophilic bentonite (BE) dispersed directly in the acrylic resin; (ii) oMMT (C25A, C93A, OMBE) dispersed in ethanol and then in the acrylic resin; and (iii) oMMT (C10A, C93A) dispersed in acetone and then in the polyurethane resin emulsion.
The first step of the sample preparation involved the dispersion and exfoliation of the clay nanolayers in an organic solvent or water. A clay suspension was prepared by adding the layered clay particles into the suspension medium, e.g., 0.5 g of untreated bentonite was added to 50 mL of deionized water. To achieve dispersion of the stacked layers, the suspension was subjected to high-shear mixing for 30 min in a batch homogenizer, Polytron PT 45/80 (Kinematica, Inc.) at 4000 rpm. The suspension was left to equilibrate for 24 h and was then sonicated for 30 min with an ultrasonic Sonotrode 400 W (Hielscher[R]) operating at 35 kHz. Similarly, for the organomodified clays, 0.5 g of these clays was added in 50 mL of acetone or ethyl alcohol for the preparation of the initial dispersion, which was then sheared and sonicated as above.
At the second step, the required amount of the exfoliated clay suspension was mixed into the aqueous matrix emulsion. For a sample containing 1.2 wt% C10A, 0.24 g of the organomodified clay was added to 30 mL of waterborne PU resin. The mixture was once again stirred and sonicated for 30 min. Then, 5 mL of the crosslinker was added, as the volumetric ratio of pure polymer resin to crosslinking agent should be 7/3. Taking into consideration the wt% of the modifier on the clay surface (Table 3) and the resin content in the waterborne suspension, this produces 17 mL of nanocomposite after drying/hardening, containing 0.146 g pure clay. The volume fraction of this sample is 0.64 vol% as is determined by applying equations 3 and 4.
Several nanocomposite coatings with varying clay contents were prepared using this method. The final nanocomposite dispersions were smeared on substrates made from glass microfiber filters, until the filters were completely covered by the mixture. The crosslinking reaction in the acrylic resin took place upon drying at room temperature and resulted in fully cured coatings. The polyurethane-based films were let to dry at room temperature for 24 h and were, then, conditioned in a drying oven for a total period of 48 h. The crosslinking reactions took place during this thermal cure. The temperature in the oven was increased gradually from ambient to 80[degrees]C to avoid the rapid evaporation of the solvent and ensure that no bubbles or pores were created on the surface of the film. It was kept below the glass transition temperature of the PU resin. After hardening, a compact (nonporous) membrane was produced, of total thickness, d, between 0.5 and 1 mm. The material occupied all the space on and between the fibers of the substrate filters. The average thickness, d, of the film (the cured coating itself, excluding the substrate) was calculated as
d = [m.sub.s] - [m.sub.f]/[[rho].sub.comp]A, (1)
where [m.sub.s] is the mass of the sample, including the substrate filter, [m.sub.f] is the mass of the substrate, [[rho].sub.comp] is the density of the nanocomposite, and A is the nominal area of the sample (directly exposed to the gas penetrants).
X-ray diffraction patterns were recorded on a D8 Advance, Bruker AXS, diffractometer with a CuK[alpha] anode. The spectra were collected in the region between 2[degrees] and 15[degrees]. The scanning step was 0.02[degrees], the time step was 0.2 s, and the X-ray wavelength was [lambda] = 1.5405 [Angstrom]. The determination of the plane spacing, [d.sub.(001)], was based on Bragg's equation:
n[lambda] = [2d.sub.(001)] sin [theta]. (2)
The volume fraction of the samples, [phi], was calculated from the mass fraction used, w, and the nanocomposite density, [[rho].sub.comp]:
[MATHEMATICAL EXPRESSION NOT REPRODUCIBLE IN ASCII]. (3)
[phi] = [[rho].sub.comp]/[[rho].sub.particles] w. (4)
The clay mass fractions of the samples were verified by TGA measurements. A Perkin-Elmer Diamond[R] TG/ DTA Instrument was used over the temperature range from 100 to 950[degrees]C, with a heating rate of 50[degrees]C [min.sup.-1] and sampling rate of 0.5 points [s.sup.-1]. The fractional mass of the residue (ash) left after completely burning the resin was compared with the value of w used to prepare the sample. The neat polymer resins were used to calibrate the instrument, as they returned values of zero residue.
DSC measurements were performed with a PL-DSC, Thorn Scientific Services Ltd, to determine the glass transition temperature of the resins. The samples were first heated to 75[degrees]C and maintained at this temperature for 1 min. They were, then, heated from 75 to 200[degrees]C, at a heating rate of 10[degrees]C [min.sup.-1]. The temperature was maintained at 200[degrees]C for 2 min and the samples were cooled down to 75[degrees]C at the same rate of 10[degrees]C [min.sup.-1]. The same cycle was performed once again for each sample.
Gas transport measurements were conducted in a permeation cell, (4) where the permeability, P, the diffusion coefficient, D, and the solubility coefficient, S, of the coating were measured. The cell has two compartments separated by the nanocomposite membrane. The test gas (C[O.sub.2]) is introduced in the left chamber and kept at constant pressure, [p.sub.0]. The concentration of the gas in the right compartment, [C.sub.d](t), is monitored to estimate the amount that permeated through the membrane. Prior to every measurement, the cell was purged with helium to remove any possible residue of C[O.sub.2]. The permeation measurements were conducted at 1 atm, 50% relative humidity and temperatures of 14[degrees]C for the acrylic resin and 21[degrees]C for the polyurethane resin.
The details and the theory of these measurements are given by Choudalakis et al. (20) When the [c.sub.d](t) curve is made, the diffusion coefficient, D, is estimated from the time lag, [T.sub.L], needed by the curve to reach a steady slope:
D = [d.sup.2 ]/6[t.sub.L] (5)
and the permeability, P, and the solubility, S = P/D, from this slope:
slope = ApP/Vd (6)
with A = 23.76 [cm.sup.2], the surface area of the membrane and d its thickness. The solubility and the diffusion coefficient are measured, thus, independently. (8)
FTIR spectra were also employed to detect the structural differences between the clay powders and the composites. KBr pellets were used in conjunction with the organoclays (1 mg of sample for 100 mg of KBr). The FTIR spectrometer acquired the IR spectra with a spectral resolution (and step width) of 2 [cm.sup.-1] in the 4000-400 [cm.sup.-1] and a total of 1801 data points per scan.
Transmission electron micrographs of some samples were taken at the Department of Biology of the University of Crete. These images were used solely for the qualitative examination of the dispersion and the orientation of the clay particles in the final composite.
Results and discussion
The structure of the nanocomposites
The TGA residue measurements confirmed that the amount of clay loading in the final nanocomposite samples was in agreement with the values used for their preparation, within the accuracy limits of the instrument. Table 2 lists the prepared samples and their actual clay volume fractions (inorganic content).
The DSC experiments, conducted to determine the glass transition temperature of the waterborne resins, gave a [T.sub.g] of about 100[degrees]C for the Sx resin. The PU resin showed a [T.sub.g] in the range of 120-130[degrees]C. Determining the glass transition temperature allowed us to perform the thermal cure of the polyurethane resin below the [T.sub.g], avoiding, thus, the formation of defects on the surface of the coating.
A typical TEM image is shown in Fig. 2 to provide visual information regarding the distribution of bentonite particles into the acrylic resin. The dark lines and shadows that can be seen in the brighter background of the matrix are the sheets of MMT clay, which are cut by the plane of the image. The micrograph indicates the state of exfoliation for the 0.91 vol% bentonite/acrylic resin nanocomposite. Numerous individual layers can be observed in the micrograph verifying that the clay is well dispersed in the matrix. A region of agglomerated layers can also be identified.
The transmission electron micrographs have the form of Supplementary Information to the XRD analysis. Both techniques are useful for the understanding of the distribution of the clay particles into the organic matrix and the analysis of the overall nanostructure.
Figure 3 and Table 3 show the (001) peaks of the XR diffractograms of the clays. MMT in its pristine state shows this peak at 2[theta] = 6.4[degrees]. The corresponding peak for the organomodified clays lies at smaller angles, around 3[degrees] to 5[degrees]. This indicates that the organic modifier pushes the clay follicles apart to form slightly intercalated structures. The increase in the intergallery spacing strongly depends on the type and conformation of the alkyl chains of the ammonium ion. Apparently, as the surface hydrophobicity of the clay increases, the d-spacing increases. Thus, the surface-treated clays may be compatible with a variety of organic systems.
When hydrophilic bentonite is dispersed directly in the waterborne acrylic resin to form the coating, exfoliation is achieved for up to 1 vol% loading. (8) The bentonite powder shows a primary, (001), peak at around 6.5[degrees]. For low mass fractions of BT this peak is absent in the curves of the composites, indicating that the nanoclay particles are well dispersed and exfoliated in the acrylic matrix. However, the peak reappears at higher volume fractions indicating reaggregation of the nanoparticles then.
Figure 4 depicts the structure of nanocomposites with the organically modified nanoparticles dispersed in the acrylic resin. Figure 4a shows samples with different volume fractions of organomodified bentonite (OMBE). All samples seem to be well exfoliated, since no obvious diffraction peak is detected. However, a careful observation can reveal a small peak at an angle around 6.2[degrees], especially for the samples with the highest loadings, 1.23 and 1.5 vol%. As the peak of OMBE powder lies at 20 = 4.1[degrees], and the unmodified clay at 2[theta] = 6.4[degrees], this peak of the nanocomposite indicates that the ethanol, which was used as solvent during preparation, may have extracted part of the octadecylamine from the clay surface. This could make it possible for some particles to re-aggregate at high vol% fractions at gallery distances equal to those in the unmodified clay.
The other oMMTs (C25A and C93A) show similar results. At low volume fractions (Figs. 4b and 4c), the diffraction peaks disappear completely. Another weak and broad peak at around 5[degrees] for C25A and 4[degrees] for C93 forms at higher volume fractions (0.6 and 0.8 vol%), an indication of incomplete exfoliation. The ethanol seems to have an effect also on the surfactant of C25A but less on C93A. The difference may be due to the observed less than complete initial dispersion of the C93A clay in ethanol.
On the other hand, when the more hydrophobic PU polymer is employed, the use of acetone for the initial dispersion for the clay types C10A and C93A seems to produce better exfoliation in the nanocomposites. These clays are completely exfoliated in the nanocomposites at 0.32, 0.64, and 0.96 vol% (Figs. 5a and 5c). On the contrary, when ethanol is the initial dispersion medium (Figs. 5b and 5d), the C10A and C93A nanocomposites are not exfoliated in this resin, even for very low clay addition; a diffraction peak appears at 2[theta] [approximately equal to] w 6.5[degrees], close to the angle of the untreated MMT (BT). It seems, thus, that, when the PU matrix is used, acetone is a more appropriate dispersion medium for these clays than ethanol.
From the results in this section, it seems that there are three main parameters affecting the final state of dispersion in the nanocomposite: (i) the initial dispersion level of the clay in the solvent; (ii) the degree of interaction between the solvent and the modifier of the particles; and (iii) the compatibility between the modified particles and the resin. Apparently the complete satisfaction of all these requirements is difficult. The results, though, indicate that acetone is a suitable clay dispersion medium for the PU matrix, whereas ethyl alcohol is better for the acrylic resin.
Table 4 lists some characteristic IR absorption bands for the clay particles acquired in the region 4000-400 [cm.sup.-1] and their possible assignments. Figure 6 is provided to enable the comparison between the IR bands of the untreated and the organomodified MMTs and the nanocomposites. Figure 6b is a magnified result which demonstrates the additional peaks for oMMT due to the presence of the organic modifiers, in the region 2950-2800 [cm.sup.-1].
The BT powder and the oMMT clays show absorption bands in several wavenumber ranges. The polyurethane matrix shows strong absorption in the 800-1600 [cm.sup.-1] range a.o. When the clay is added to the matrix the IR spectrum that is obtained shows the bands both of the clay and the matrix. Thus, some parts of the spectrum are very crowded with overlapping peaks. The bands at the remaining ranges are more appropriate for the examination of changes (shifting of peaks) that can be used to study the state of dispersion in the nanocomposites.
The characteristic peak that is seen at 1032 [cm.sup.-1] for the untreated MMT is shifted toward higher wavenumbers for the organomodified clays: 1036 [cm.sup.-1] for OMBE, 1046 [cm.sup.-1] for C10A, 1048 [cm.sup.-1] for C25A, and 1052 [cm.sup.-1] for C93A (Fig. 6c). Similar shifts are observed for the peaks at 3626 [cm.sup.-1] (pure MMT), 3628 (C10A and C25A), and 3636 [cm.sup.-1] (OMBE and C93A).
This shift follows a well-defined trend that can be related to differences in the cation exchange capacity (CEC, layer charge) of the surfactants used. (21,22) The CEC represents the surface charge of the clay and determines the amount of cationic surfactant that can be intercalated into the galleries by ion exchange. The CEC of the clays is shown in Table 3: as CEC decreases the absorption peaks shift to higher wavenumbers. Further, comparing this shift with Fig. 3, one can see that the trend follows the observed increase of the interlayer distance seen via XRD for these samples (also in Table 3). Thus, these IR band shifts indicate changes in the intergallery distance.
The peak of the montmorillonite around 1032 [cm.sup.-1] overlaps with peaks of the PU matrix in that range and it is not suitable to be used for the characterization of the nanocomposites. The most useful range of IR absorption bands for the latter may be between 600 and 400 [cm.sup.-1] (Fig. 6d). The peaks for the nanocomposite samples at that range seem to be slightly shifted relative to the ones of the (intercalated by the modifier) clay powders (2-6 [cm.sup.-1]). However, this range of wavenumbers are very close to the edge of the measuring range of the instrument, while the shifts are marginally within its resolution limits. These results can be considered only as indirect indication of exfoliation of the clay but they may give limited support to other more accurate ways used to judge the dispersion level of the clay particles.
The effect of the clays on C[O.sub.2] gas diffusion
Figures 7 and 8 display the relative diffusivity in the Sx/BT, Sx/oMMT, and PU/oMMT nanocomposite coatings. The error bars reflect the fluctuations in the local thickness of the film. As expected, when the particle volume fraction increases, the diffusion coefficient decreases, e.g., by about 40% for 1.3 vol% BT (Fig. 7a). In the OMBE nanocomposites, Fig. 7b, D decreases almost linearly. The sample with 1.5 vol% clay has a value of D about 60% of that of the pure resin. The decrease does not follow the tortuous path model proposed by Nielsen. (4,6) The failure of the model in this case may be an indication that there is another significant contribution to the diffusion process, probably due to the changes in the free volume around the interfaces in the system upon addition of more clay. (7)
The diffusion behavior of C25A and C93A in Sx nanocomposite coatings (Fig. 7c) can be described well by the tortuous path model. Fitting Nielsen's equation predicts a value for the aspect ratio L/W [approximately equal to] 200 for the particles in both nanocomposites. This relatively high value verifies the good exfoliation of the samples. These clays were initially dispersed in ethanol before being added to the waterborne acrylic matrix. At higher volume fractions, the degree of exfoliation is predicted to be lower, something that is confirmed by the X-ray results. The overall reduction of the diffusion coefficient for ~0.6 vol% C25A or C93A particles is around 30%.
The relative diffusivity for the C10A and C93A clay/ PU nanocomposites is displayed in Fig. 8. The employed dispersion medium in this case was acetone. The measurements were conducted at a temperature of 21[degrees]C and relative humidity of 40-50%. The incorporation of 0.64 or 0.96 vol% of C10A results in a decrease of the diffusion coefficient in the order of 37% or 68%, respectively. Similar reduction is achieved for the C93A nanoclay, where the decrease is 34% or 72% for addition of 0.64 or 0.97 vol% clay, respectively.
On the other hand, the solubility coefficient increases much faster in the organomodified cases than for untreated MMT (Figs. 7 and 8). This is evidence that there may be an interaction/affinity between the penetrating carbon dioxide molecules and the organic modifier of the nanoparticles. Therefore, it seems that the clay/matrix compatibilizer (organic modification) may have two opposite effects: it may improve the interactions between the two components of the nanocomposite, but it may also enhance the adsorption of the gas molecules in the interfacial regions between the organic matrix and the inorganic crystalline layers.
All these results indicate that the enhancement of gas barrier properties in a nanoplatelet-reinforced matrix is affected by three parameters: (i) the aspect ratio of the inorganic particles; (ii) the possibility of extra free volume created at the clay/matrix interfaces during material preparation; and (iii) possible interactions between the organic modifiers of the clay and the permeating gas.
The extra free volume at the interfaces could be due to the constraints imposed there on the long aliphatic chains of the organic modifier. The existence of such free volume areas has not been verified by PALS measurements. (7) However, the possibility of the existence of such areas cannot be ruled out because PALS may not be able to accurately probe these areas due to their high electronic density. The problem of the possible interactions of the permeating gas and the organic modifier is specific to the gas/membrane combination and is beyond the scope of the present article.
The most important parameter for the barrier properties is the aspect ratio of the particles. (9) When a large value of particle L/D is retained during the preparation of the nanocomposite by avoiding aggregation of the follicles, then it should be expected that the diffusion of gases through the coating will be drastically reduced. Thus, it is of paramount importance that the complete exfoliation of the clay is preserved in the final composite.
The compatibility of suspension media and organomodified clays
The hydrophilic MMT has been found to produce well-exfoliated dispersions in waterborne, relatively hydrophilic (acrylic) resin emulsions up to 1 vol%. In inherently more hydrophobic resins (e.g., PU), its dispersion is less successful. In these matrices, organomodification seems to provide an improvement. Orgnanomodification of the clay surfaces is necessary when the clay is to be dispersed in organic solvent systems.
In any case, the dispersion of the organomodified clays in a waterborne resin requires the initial dispersion of the particles in a compatible solvent-based dispersion medium. The nature of solvents, therefore, and their compatibility with the clay modifier and the water could determine the quality of this dispersion and the final nanocomposite structure. (23) A usual method to study the compatibility of two substances is to compare the values of their solubility parameters. (24)
The solubility parameter, [delta], provides an estimation of the cohesive energy of a substance and is used to establish the miscibility/compatibility of the components of a mixture. In the present work, the value of [delta] was used as a criterion for the suitability of a solvent for the development of the nanocomposite.
The value of [delta] is estimated by Fedors Group Contribution Theory. (25) This method adds the cohesive energy, [E.sub.coh,i], and molar volume, [V.sub.m,i], contributions of each functional group and each fragment of the structure to find the value of [delta] of the molecule:
[delta] = [([[summation].sub.i] [E.sub.coh,I]/[[summation].sub.i] [V.sub.mi,i]).sup.1/2]. (7)
The data for the functional groups involved in the organomodifiers are presented in Table 5. Table 6 lists the solubility parameters for the organic modifiers of the MMT nanoclay and the solvents used in the present work.
The solubility parameters of the organomodifiers have a value around 17 [MPa.sup.1/2]. It is evident, from Table 6, that among the solvents that were tried, this value is closer to the solubility parameters of o-xylene and cyclohexane. However, these solvents are not compatible with water. Thus, they cannot be used as dispersion medium of the clay when waterborne resins are used. Acetone and ethanol are both miscible with water and have slightly higher values for [delta]. The values of [delta] of the modifiers are closer to that of acetone than that of ethanol. Thus, the dispersion should be easier when the former solvent is used as initial suspension medium than the latter. The exfoliation is carried through to the composite when the initial suspension is added to the waterborne resin emulsion.
When acrylic resin is used, both acetone and ethanol give rather good results (exfoliated clay in the nanocomposite). Acetone seems, indeed, to have a slight advantage over ethanol as the initial dispersion medium, and more clay can be added before reaggregation reappears in the composite. When waterborne polyurethane is used as the matrix, then acetone is much better than ethanol in producing exfoliated composites. This points out the importance of maintaining a relative compatibility between all different components of the nanosystem at all stages of its preparation: the clay and/or clay organomodifiers, the suspension medium (solvent) for the initial dispersion, and the resin with its resultant presence of water.
It seems that high barrier properties result from a well exfoliated and dispersed clay in the matrix. In order to achieve the dispersion of the organically modified MMT in an aqueous resin emulsion and the formation of a well-exfoliated nanocoating, various parameters need to be taken into consideration. Mainly, these are the miscibility between the solvent and the modified clays, the miscibility between the solvent and water and the compatibility between the polymer matrix and the modified particles. Since the preparation method forced the dispersion of the nanolayers into a solvent, prior to their dispersion into the polymer matrix, it is reasonable to claim that the former process (rather than the latter) controls and determines, to some degree, the final properties of the film. Therefore, it is useful to evaluate and determine the physical parameters that dominate the clay dispersion process in the solvents used. The analysis according to group contribution theory suggests that the solubility parameters of the clay surfactant should be close in value to the ones of the solvents used as the initial dispersion medium.
However, the above miscibility criterion alone does not guarantee the initial interlayer swelling of the clay in the solvent. If such a swelling does not occur, then the exfoliation of the clay in the final dispersion may be incomplete. The swelling factor (miscibility) and the interlayer swelling are two independent parameters. The former is determined by comparing the solubility parameters of the solvent and the modifier as described above. The latter depends on the polar character of the solvent/dispersion medium and the swelling capacity of the clay, and it is determined by the surface tension.
Nanocomposite coatings can be prepared by incorporating hydrophilic or organomodified MMT nanoclay into different aqueous resin emulsions. These coatings present improved barrier properties against gases like C[O.sub.2]. The nanoparticles create tortuous paths within the matrix, hindering the diffusion of the gas molecules through the coating. Various parameters influence the barrier properties of the coating such as the aspect ratio and the volume fraction of the inorganic nanoparticles and the type of the organic modifier present on their surface. The dispersion media used for the preparation of the composite coating and the thermodynamic interactions between the organic component and the inorganic reinforcement also influence its final barrier properties.
The correlation between the degree of dispersion (exfoliation) and the barrier properties of the coating can be verified by direct measurements of gas permeation and X-ray structural analysis. Experimental support for the structure is obtained also from IR absorption measurements.
Nanocomposite coatings with exfoliated structures can be prepared by initially dispersing the organomodified clay in solvents and then mixing this suspension with the waterborne resin emulsion. The solvent used as the initial dispersion medium for the clay should be miscible with water. When the solubility parameter of the solvent is close in value to the one of the clay surfactant, then the state of exfoliation in the final composite is sufficient for up to 2 vol% particle content. Untreated clay can also be dispersed in relatively more hydrophilic waterborne resins, such as the acrylic resin used here. The barrier properties of all these coatings are significantly improved.
The type of modified clay should be chosen according to its initial d-spacing; the initial distance between the layers (determined by the organic modifier) of the clay should be large to facilitate the entry of the solvent molecules into the intergallery regions, leading to a more efficient swelling of the nanoclay. Further, the temperature during drying of the coating and the curing of the thermosetting matrix should be such as to allow the timely evaporation of the solvent/water mixture and avoid bubbles or pores appearing in the bulk or on the surface of the film.
Acknowledgments This research was partially supported by the program THALIS "NAMCO" #681135, which was financed by the European Union (European Social Fund) and Greek national funds (Program E[summation][PI]A). Collaborations on the subject of gas permeation through nanocomposites have been had with Prof. Reinhard Krause-Rehberg of the Martin Luther University of Halle-Wittenberg to whom we are grateful. We would also like to thank Prof. G. Chalepakis and Dr. A. Siakouli of the University of Crete for the TEM images, and Prof. Noni Maravelaki of the School of Architecture of the Technical University of Crete for her help with the IR measurements and their interpretation.
(1.) Ray, SS, Okamoto, M, "Polymer/Layered Silicate Nanocomposites: A Review from Preparation to Processing." Prog. Polym. Sci., 28 1539-1641 (2003)
(2.) Usuki, A, Hasegawa, N, Kato, M, "Polymer-Clay Nanocomposites." Adv. Polym. Sci., 179 135-195 (2005)
(3.) Esfandiari, A, Nazokdast, H, Rashidi, AS, Yazdanshenas, ME, "Review of Polymer-Organoclay Nanocomposites." J. Appl. Sci., 8 545-561 (2008)
(4.) Choudalakis, G, Gotsis, AD, "Permeability of Polymer/Clay Nanocomposites: A Review." Eur. Polym. J., 45 967-984 (2009)
(5.) Sorrentino, A, Tortora, M, Vittoria, V, "Diffusion Behavior in Polymer-Clay Nanocomposites." J. Polym. Sci. B, 44 265-274 (2006)
(6.) Nielsen, LE, "Models for the Permeability of Filled Polymer Systems." J. Macromol. Sci. A, 1 929-942 (1967)
(7.) Choudalakis, G, Gotsis, AD, "Free Volume and Mass Transport in Polymer Nanocomposites." Curr. Opin. Colloid Interface Sci., 17 132-140 (2012)
(8.) Choudalakis, G, Gotsis, AD, "Morphology and Gas Transport Properties of Acrylic Resin/Bentonite Nanocomposite Coatings." Prog. Org. Coat., 77 845-852 (2014)
(9.) Choudalakis, G, Kalo, H, Breu, J, Gotsis, AD, "Gas Barrier Properties in Polymer Nanocomposite Coatings Containing Li-Hectorite Clays." J. Appl. Polym. Sci. 2014. doi:10.1002/ APP.40805
(10.) Nobel, ML, Mendes, E, Picken, SJ, "Enhanced Properties of Innovative Laponite-Filled Waterborne Acrylic Resin Dispersions." J. Appl. Polym. Sci., 103 687-697 (2007)
(11.) Patel, MJ, Gundabala, VR, Routh, AF, "Modeling Film F0ormation of Polymer-Clay Nanocomposite Particles." Langmuir, 26 3962-3971 (2010)
(12.) Strawhecker, KE, Manias, E, "Structure and Properties of Poly(vinyl alcohol)/Na Montmorillonite Nanocomposites." Chem. Mater., 12 2943-2949 (2000)
(13.) Valadares, LF, Linares, EM, Braganca, FC, Galembeck, F, "Electrostatic Adhesion of Nanosized Particles: The Cohesive Role of Water." J. Phys. Chem. C, 112 8534-8544 (2008)
(14.) Vinjamur, M, Caincross, RA, "Non-fickian Nonisothermal Model for Drying of Polymer Coatings." AIChE J., 48 2444-2458 (2002)
(15.) Wang, T, Keddie, JL, "Design and Fabrication of Colloidal Polymer Nanocomposites." Adv. Colloid. Inter. Sci., 147-148 319-332 (2009)
(16.) Wicks, ZW, Jones, FN, Pappas, SP, Wicks, DA, Organic Coatings, Science and Technology, 3rd ed. Wiley, New York, 2007
(17.) Sewell, G, "Importance and Measurement of Minimum Film-Forming Temperature." Pigment Resin Technol., 27 173-174 (1998)
(18.) Stefanis, Em, Panayiotou, C, "Prediction of Hansen Solubility Parameters with a New Group-Contribution Method." Int. J. Thermophys., 29 568-585 (2008)
(19.) Nobel, L, "Water-Borne Nanocomposite Coatings." PhD Thesis, Delft University of Technology, 2600 AA Delft, the Netherlands, 2007
(20.) Choudalakis, G, Gotsis, AD, Schut, H, Picken, SJ, "The Free Volume in an Acrylic Resin/Laponite Nanocomposite Coatings." Eur. Polym. J., 47 264-272 (2011)
(21.) Nahin, PG, "Infrared Analysis of Clays and Related Minerals." Clays Clay Technol., 169 112-118 (1952)
(22.) Karakassides, MA, Petridis, D, Gournis, D, "An Infrared Reflectance Study of Si-O Vibrations in Thermally Treated Alkali Saturated Montmorillonites." Clay Miner., 34 429-438 (1999)
(23.) Burgentzle, D, Duchet, J, Gerard, JF, Jupin, A, Fillon, B, "Solvent-Based Nanocomposite Coatings. Dispersion of Organophilic Montmorillonite in Organic Solvents." J. Colloid. Interface Sci., 278 26-39 (2004)
(24.) Ho, DL, Glinka, CJ, "Effects of Solvent Solubility Parameters on Organoclay Dispersions." Chem. Mater., 15 1309-1312 (2003)
(25.) Fedors, RF, "A Method for Estimating Both the Solubility Parameters and Molar Volumes of Liquids." Polym. Eng. Sci., 14 147-154 (1974)
M. Stratigaki, G. Choudalakis, A. D. Gotsis ([mail])
Technical University of Crete, 73100 Hania, Greece
Table 1: Properties of the organic modifier for the MMT clay types considered in this study Trade name Modifier Chemical structure C10A (2MBHT) Dimethyl, benzyl, hydrogenated [FORMULA NOT tallow quaternary ammonium REPRODUCIBLE IN ASCII] C93A (M2HT) Methyl, bis-hydrogenated tallow [FORMULA NOT quaternary ammonium REPRODUCIBLE IN ASCII] C25A (2MHTL8) Dimethyl, hydrogenated tallow, [FORMULA NOT 2-ethylhexyl ammonium REPRODUCIBLE IN ASCII] OMBE Octadecylamine [FORMULA NOT REPRODUCIBLE IN ASCII] The hydrogenated tallow (HT) had a composition of 6% [C.sub.18], 30% [C.sub.16], and 5% [C.sub.14] Table 2: List of samples Code Modifier Resin Solvent Volume fractions (%) BT None Sx None 0.3, 0.35, 0.61, 0.91, 1.3, 2.1 OMBE Octadecylamine Sx Ethanol 0.5, 0.75, 1.00, 1.23, 1.50 C93A C93A (a) Sx Ethanol 0.10, 0.25, 0.39, 0.60, 0.80 C25A C25A (a) Sx Ethanol 0.16, 0.32, 0.48, 0.63, 0.74 C10A C10A (a) PU Acetone 0.32, 0.64, 0.96 C10A (a) PU Ethanol 0.32, 0.64, 0.96 C93A C93A (a) PU Acetone 0.32, 0.64, 0.97 C93A (a) PU Ethanol 0.32, 0.64, 0,97 The fourth column lists the solvent used together with the water as dispersion medium. The volume fractions correspond to neat clay (without the modifier) and were verified by TGA (a) See Table 1 Table 3: Diffraction peak values calculated from X-ray data graphs and intergallery spacing of untreated and organically modified clays Clay type MMT C10A C25A C93A OMBE Modifier (wt%) 0 39 34 37.5 35 Density (g [cm.sup.-3]) 2.60 1.90 1.87 1.88 1.71 Diffraction peak (2[theta], 6.36 4.59 3.59 3.40 4.10 [omicron]) d-Spacing ([Angstrom]) 13.9 19.2 24.6 26.0 21.5 CEC (meq/100 g) (a) -- 125 95 90 (a) The values of the cation exchange capacity (CEC) were given by the supplier Table 4: Positions ([cm.sup.-1]) of some IR vibration bands observed in the range of 4000-400 [cm.sup.-1] Hydrophilic: Organomodified MMT Bond motion MMT OMBE C10A C25A C93A 3626 3636 3628 3628 3636 Al in octahedra -- 2920 2922 2926 2924 nonsymmetric stretching C[H.sub.2] -- 2850 2850 2852 2850 symmetric stretching C[H.sub.2] -- 1468 1470 1470 1470 C-C[H.sub.2] bending 1032 1036 1046 1048 1052 Si-O stretching 916 916 918 918 918 Al-Al-OH bending 518 522 522 522 522 Al-O-Si, Si-O-Si, Si-O 464 464 464 462 464 Al-O-Si, Si-O-Si, Si-O Table 5: Group contributions to the cohesive energy and molar volume of functional groups used to estimate the solubility parameters for the organomodifiers of MMT nanoclay (25) Group [E.sub.coh] [V.sub.m] ([cm.sup.3] (J [mol.sup.-1]) [mol.sup.-1]) -C[H.sub.3] 4707 33.5 -C[H.sub.2]- 4937 16.1 -OH 29,790 10 Phenyl 31,924 71.4 Nitrogen 4184 9 Table 6: Solubility parameters of the organic modifiers and solvents Surfactant C10A C93A C25A OMBE Solubility parameter 17.1 17.1 16.5 17.6 ([Mpa.sup.1/2]) Solvent Ethyl Acetone o-Xylene Cyclohexane alcohol Solubility parameter 26.6 20.1 18 16.8 ([Mpa.sup.1/2])
Please note: Some tables or figures were omitted from this article.
|Printer friendly Cite/link Email Feedback|
|Author:||Stratigaki, M.; Choudalakis, G.; Gotsis, A.D.|
|Publication:||Journal of Coatings Technology and Research|
|Date:||Nov 1, 2014|
|Previous Article:||Electrochemical behavior of hydrophobic silane-zeolite coatings for corrosion protection of aluminum substrate.|
|Next Article:||Surface properties of alicyclic epoxy coatings modified with reactive copolymer containing fluorinated side chains.|